Interview with Paul S. Chan, MD, MSc

Paul S. Chan, MD, MSc is Associate Professor of Medicine at the University of Missouri, Kansas City and a clinical cardiologist with Saint Luke's Mid America Heart and Vascular Institute.

In this interview with Angioplasty.Org, Dr. Chan discusses his team's recent study, "Appropriateness of Percutaneous Coronary Intervention"¹ which was published in the Journal of the American Medical Association (JAMA) on July 5, 2011. Dr. Chan gives background on the study itself, the import of its findings and his concerns over the way the study may be interpreted. Dr. Chan served as lead author of a group of a dozen leading cardiologists to issue this first report on the issue of PCI appropriateness since the multi-disciplinary 2009 PCI Appropriate Use Criteria (AUC)² were published -- it will no doubt serve as the benchmark for further exploration of the appropriate use of stents and angioplasty in the United States.

Paul S. Chan, MD, MSc

Q: Tell us about the genesis of your study of PCI Appropriateness¹ and who was involved in writing it.
Dr. Chan: There have been several studies historically that have looked at the appropriateness of angioplasty using different criteria, but oftentimes they would adjudicate based on criteria that were author-based, or that were limited in terms of collection of information of symptoms and non-invasive test results. When the Appropriate Use Criteria (AUC) for coronary revascularization, sponsored by the national societies, such as SCAI and ACC, were developed and were subsequently published in January of 2009, we as a group decided that it would be important and informative to get a better sense of the contemporary rates of PCI appropriateness, using these more standardized criteria.

The group was really developed to have a diversity of players, so it wouldn't be driven by interventionalists who may have a particular angle to the story or, for that matter, non-interventionalists: we wanted a balanced perspective. For example, I'm a clinical cardiologist. So we engaged people who had been involved in the development of the criteria or in the leadership of the ACC and SCAI to collectively draft this document. And we thought this would be particularly helpful in the context of more recent reports of individuals in the press that seemed to highlight maybe blatant overuse of angioplasty procedures.

Q: These appropriateness criteria were written in 2009. Two years earlier, when the COURAGE study came out, it seemed that there was a sort of split between cardiologists: "Should we use drugs or stents? There’s no difference. Angioplasty is being radically overused,” and so on. So were you surprised by the numbers in your study, compared to what had been estimated previously?

"There were predictions after COURAGE as to whether or not angioplasties were in fact 50% inappropriate, whereas we found a much lower rate in the elective setting, of about 12%."

Dr. Chan: Sure. If the Appropriate Use Criteria were based on the COURAGE trial itself only, then I think we probably would have different numbers. The COURAGE trial was certainly in the mindset of the technical panel that developed the ratings for each of the clinical scenarios, but it was not the sole criterion, although medical therapy certainly plays an important role in understanding whether or not procedural appropriateness would be appropriate in certain circumstances. There were predictions after COURAGE as to whether or not angioplasties were in fact 50% inappropriate, whereas we found a much lower rate in the elective setting, of about 12%. If the appropriate use criteria were just developed based on the COURAGE criteria, I think certainly that the quote-unquote "inappropriate" or whatever you want to call it rate would be higher, but that's an important distinction, to make sure that people understand that these criteria are not an extension of the COURAGE trial in and of itself, that they really reflect the totality of both clinical trial evidence and expert opinion as to when angioplasty might be beneficial to patients.

Q: Speaking of COURAGE – that was about stable elective patients only, not acute or emergency situations, but the headlines in the mass media shouted “Stents Don’t Work” and didn’t really discern the very important difference between acute and non-acute scenarios, doing a disservice to patients.

Dr. Chan: Yeah. I think, the clarity, the distinction between acute and non-acute is a difficult one for non-medical people to understand, and certainly lay people, and I think we probably should be describing the difference as "heart attacks” and “non-heart-attack-related syndromes." Because I think people understand what a heart attack is, but "acute" is less familiar to the majority of not only journalists, but also people who will be hearing about our study.

"You can see the press potentially saying, '50% of elective angioplasties are not appropriate….' And that IS a misinterpretation!"

We also worry about whether or not the focus will be less on the “inappropriate” rate, but more on the “not appropriate” rate (Ed. Note: “not appropriate” means “inappropriate plus uncertain”) --- meaning you can see the press potentially saying "50% of elective angioplasties are not appropriate…." And that IS a misinterpretation of the terms!

Unfortunately, the terms that are used in the criteria are very difficult. They have lay interpretations; people understand "inappropriate" and "appropriate" in very different ways, and I'm certain, for that matter, that people might misinterpret as not having any benefit altogether, whereas it's really that the body of clinical evidence is just not overwhelming enough to suggest that there's conclusive benefit, but there is likely to be some benefit, especially in that large number of patients with “uncertain” procedures. So that's where I think our worry about where the press might take this will be.

What I think we want to really focus on is the fact that the “appropriate” procedures really suggest a definitive or probable benefit, “uncertain” procedures really suggest a possible benefit, and “inappropriate” procedures suggest that there's unlikely to be benefit. That doesn't mean that there's no circumstance when an inappropriate patient has no benefit; it means that, on a population average, patients who have inappropriate procedures are not going to gain as much in terms of symptom benefit or health status improvement as patients who had a clinical and appropriate procedure.

Q: Right after COURAGE, most of the interventional cardiologists I spoke with said that the COURAGE trial, which looked at stable patients only, would not change their practice since the majority of angioplasties they did were in acute patients. And here we are four years later, and they were right -- that's pretty much what the results of your study show: 71% of angioplasties were done for acute cases. Did that surprise you as well?
Dr. Chan: Yeah, we were certainly not anticipating that a little bit more than two-thirds of procedures that we mapped would be acute. Part of it I think is the fact that we had to exclude a number of cases, a hundred thousand or so, primarily in the elective setting, because we did not have information on stress testing -- either those patients went straight to angiography without stress testing beforehand, or there was no documentation of what the stress test result was. So, if we had actually included those, we probably would have seen a closer to 50/50 split of acute and non-acute procedures.

As to whether the rates of inappropriate and appropriate and uncertain PCI really reflect the totality of the procedures that were elective in the registry, unfortunately, we just don't have enough information from those excluded procedures to better know whether or not they would have been appropriate. We did look at it, but we didn't really describe in the paper, and I suspect that some people will describe this subsequently.

Q: Besides stress testing, how did you factor in other imaging modalities, for example Cardiac CT, Intravascular Ultrasound (IVUS) and Fractional Flow Reserve (FFR)?
Dr. Chan: We actually excluded patients that had CT angio, or just coronary calcium testing, because they didn't really have an ischemia assessment, and we couldn't map the patients unless they had some sort of ischemia evaluation. As for FFR -- the clinical scenarios in the Appropriate Use Criteria did ask about patients who had less than 70% stenosis and the use of FFR and IVUS, but it did not ask for FFR data in patients who had 70% or greater stenoses, so those patients would have been excluded because they didn't have an ischemia evaluation.

Q: Interesting since, if an FFR was done during the diagnostic cath and PCI was then performed, the procedure would have been “appropriate” because, according to the FAME study, FFR is an important tool for evaluating whether a blockage is causing ischemia or not.
Dr. Chan: Ultimately, we may need the Appropriate Use Criteria Version 2.0 that incorporates FFR in patients who didn't have a prior stress testing, but had a 70% stenosis and then subsequently an FFR during the procedure. We were aware of the FAME results, and we certainly could have mapped those procedures had they been assigned a rating in the Appropriate Use Criteria, but we did not feel that we should've assigned those ratings ourselves.

Q: About the diagnostic pathway to the cath lab, there's been a lot of controversy about patients who do not get stress testing first, but they have symptoms, or the physician just says "go get an angio and, if there’s a blockage, we’ll open it up.” This ad-hoc angioplasty is a big change from the early days when you did a diagnostic cath, stopped, looked at it, discussed with the patient, perhaps with a surgeon, and waited a few days or a week or a month before going ahead with the intervention.

Dr. Chan: Yeah -- I don't know what the media and the lay press will do with our paper and how they'll interpret it but, if there are huge concerns about rates of inappropriate procedures even in the elective settings, the real question in terms of quality improvement can't just lie at the feet of the interventionalists. Because oftentimes they're referred a lot of these patients, for varying reasons, and they don't know these patients, by and large. So their job has been, historically, to perform the angiogram and then proceed to PCI if it seems reasonable. And sometimes the rationale for going to angiography may be very well-based: the patients may have severe ischemia.

"The real question in terms of quality improvement can't just lie at the feet of the interventionalists. We need to engage the referral physician base...we should really get patients involved in the decision-making process itself, and really get general cardiologists to feel like they're part of the process of treatment."

But I think there is a role for patients who go for angiography to see what the coronary anatomy is, to ascertain that there isn't a high risk anatomy, for example, left main disease or three vessel disease. If it's one or two vessel disease and the patient was not severely symptomatic, no medical therapy prior to that, with an ischemia evaluation that was low to intermediate, one could make the argument that the patient should be given a trial of medical therapy to see whether or not the symptoms stabilized before proceeding to coronary angioplasty. That facilitates a conversation with the patient of what was found. It allows for a discussion of what the likelihood of clinical benefit would be for that patient, depending upon what the symptoms and ischemia risk had been, and it gives them an opportunity to see if the ischemia risk or symptom burden is modified by being placed on medications to begin with. Ultimately, it may be better for the patient in the sense that, if the symptoms are improved with just being on medications, they could forgo a procedure that might cost them 20 or 30 percent of the co-pay, and potential bleeding risk from being on dual antiplatelet therapy. That doesn't mean that angioplasty is not beneficial to patients in all circumstances; it just means that we may need to not be as robotic or automatic in just proceeding to angioplasty.

Q: You talk about a “discussion”. Do you think this should be more of a collaborative decision between patient and physician?
Dr. Chan: Yes…and it takes the onus off the interventionalist, because it engages the referral physician, the general cardiologist, the internist, in having that discussion with the patient. So the interventionalist does not always open up every lesion that is seen, simply because he can, but ensures that this is something that ultimately can help patients feel better. We know for the vast majority of the elective setting patients, where there's no high risk coronary anatomy, the only benefit really to be had is symptom relief, health status improvement. If the patient is not severely symptomatic and those symptoms can be avoided by being on maximal medical therapy, then that might be the right way to go for patients. This is really the type of patient that we would think of in the COURAGE study.

But I think in order for this to really be realizable, we need a multi-prong approach, and it can't just be in the cath lab where that quality of improvement occurs. We need to also engage the referral physician base to better understand what the role of angioplasty and diagnostic angiography is, and how we should really get patients involved in the decision-making process itself, and really get general cardiologists to feel like they're part of the process of treatment.

"The sub-specialty of cardiology should really be applauded in taking the leadership in looking at these hard questions.... Whether it's ICDs or angioplasty or stress testing. It really starts the conversation of how to improve quality...and that is a remarkable effort, knowing that sometimes it may lead to some revenue decreases for members of the field itself."

Q: Final thoughts on what this study means for interventional cardiology?
Dr. Chan: One of the things we should remember all along is that the sub-specialty of cardiology should really be applauded in taking the leadership in looking at these hard questions, and in doing self-reflection as to quality internally. I can think of no other sub-specialty that is doing it at the level that cardiology has been doing over the last decade, whether it's ICDs or angioplasty or stress testing, they're developing appropriate use criteria and assessing rates of procedural appropriateness. It really starts the conversation of how to improve quality, how do we address issues of potential overuse, and ultimately, how can we improve the care that we deliver to patients. And that in itself is a remarkable effort, knowing that sometimes it may lead to some revenue decreases for members of the field itself.

Having said that, I think moving forward, we need to think in a novel way about how we can develop both real-time and non-real-time decision tools that can guide clinicians who either perform the procedure or who order the procedure to better understand that some procedures are not going to likely benefit patients. When a patient is referred to diagnostic angiography, we already know their symptoms. We already know their ischemia test results, if they had one. We already know whether or not they're on medical therapy. And so, you can envision physician tools, either at the point of referral or in the cath lab. Based on coronary anatomy, you already know whether or not a procedure would be considered appropriate, uncertain, or inappropriate by the Appropriate Use Criteria. And having that front and center, for both referring physicians and/or the interventionalists, may help better guide understanding of whether or not we're utilizing resources well, and what the threshold, perhaps, of even treating or sending these patients to diagnostic angiography should be.

Ultimately, it'll require a commitment by both hospitals and physicians to take on this effort, to really decrease procedures that may be truly inappropriate, and therefore ensure that patients are not harmed. What we think the most important use of the Appropriate Use Criteria will be, really, is to better understand whether or not one hospital’s rate of inappropriate procedures are high and is an outlier relative to other hospitals. And if a hospital finds itself in that situation, it may be worthy for them to really better understand what's going on, because it's unlikely that a hospital, if the national average in a hospital for inappropriate rates for elective settings is about 11% and one hospital is 25, 30%, it's unlikely that it's really just differences in patient mix, but it may be that there's just potential overuse in those settings.

Q: And the fix for that would be?
Dr. Chan: I think what the NCDR has been doing, starting in May, is now providing on a quarterly basis, the rates of appropriate, uncertain and inappropriate procedures for acute and non-acute indications. It also provides a line-by-line item list of all the inappropriate cases for the last quarter at that hospital. Hospitals should then go back to the data themselves and look at who those patients were. The report now provides the actual appropriate use criteria indication, to which these patients were mapped, and to better understand whether or not there's a pattern of quote-unquote “inappropriate” cases at that hospital. And certainly the physicians or hospitals may disagree with the appropriate use criteria, but certainly, if other hospitals are performing rates that are much lower, it's just that that hospital's performance may be somewhat of an outlier, at least, for certain indications for angioplasty.

¹ Appropriateness of Percutaneous Coronary Intervention ; JAMA. 2011;306(1):53-61.doi:10.1001/jama.2011.91

² ACCF/SCAI/STS/AATS/AHA/ASNC 2009 Appropriateness Criteria for Coronary Revascularization; J Am Coll Cardiol 2009;53:530 –53

This interview was conducted in July 2011 by Burt Cohen of Angioplasty.Org.